• Built a Kafka-based distributed execution engine that improved throughput by up to 100x, with fault-tolerant recovery and horizontal scalability.
• Improved dashboard performance and stability, achieving 20x faster load times.
• Built an AI red-teaming orchestrator for adversarial testing of APIs and AI systems, supporting customer-facing security workflows and enterprise evaluations.
• Designed and implemented a fine-grained RBAC and access management system across products, including dynamic roles and scoped permissions.
• Built backend systems for AI traffic monitoring, policy guardrails, API dependency graph construction, and MCP-integrated AI querying.

Designing backend services and internal platforms with a focus on performance, correctness, and maintainability.
I enjoy working on APIs, authorization, async workflows, data-heavy systems, and production hardening.

Interested in systems that need to behave well under load — queues, caching, retries, concurrency, consistency, partitioning, and fault tolerance.
I especially like problems around execution engines, auth systems, and event-driven workflows.

Building AI-powered workflows using LangChain, LangGraph, MCP servers, retrieval pipelines, and orchestration layers.
I’m particularly interested in making LLM systems reliable, observable, and useful in production.

Built a system that ingests OpenAPI / Swagger specs to construct dependency graphs and enable dependency-aware API replay, including AI-assisted reasoning for inferring missing relationships between APIs.

Designed and implemented a scalable Role-Based Access Control system with fine-grained permissions across products, including dynamic role handling and enforcement layers across services.

Developed an orchestrator for simulating adversarial attacks on APIs and AI systems, generating actionable reports and remediation workflows for customer-facing security use cases.

Worked on real-time monitoring and policy enforcement for AI traffic and agent workflows, including synchronous compliance APIs and security-oriented guardrails.

Designed backend architecture for an AI chatbot integrated with MCP-based system querying and internal product workflows.

• Backend architecture for scalable applications
• High-level and low-level system design
• Distributed systems and concurrency patterns
• AI application architecture
• Search, retrieval, reranking, and orchestration for LLM products